Attack detection and prevention in smart water networks

This research theme aims to utilise data intelligence and security to improve the cyber-physical systems of smart water networks. 

Researchers 

• Prof Wenyan Wu
• Haitham Mahmound
• Prof Mohamed Gaber
• Dr Leo Wang
• Dr Michael Romano (UU)
• Dr Stamatia Rizou (SLG)

Research background 

Concerns about the security of water networks have been significantly growing in recent years due to the increasing reliance on smart sensing and control systems. While these systems have equipped the water infrastructure and physical systems with a new level of service and maintenance, cyber attacks can cause catastrophic disasters and can endanger public health and safety.

Bringing security measurements and innovative approaches to water applications have become one of the core demands in this research and ongoing studies have been exploring non-financial use-cases of blockchain along with rapid implementation of data intelligence algorithms in Industrial sector. On the one hand, blockchain can help achieve the security requirements of Water 4.0 by providing data immutability, integrity, and confidentiality.

Research aims

This research aims to provide utilities with the insights about the potential attacks (i.e. Device tampering and MITM) using smart sensing, AI and blockchain technologies based on the connection and configuration of SWN and their countermeasures.

Research methods

Review the potential for attacks against smart water networks and, using the case studies from the water distribution network, review the causes of these attacks and assess the impact on CPS. Investigate the countermeasures for these potential attacks along with identifying possible scenarios and consequences. Identify the security requirements, exploit the smart sensing and AI, blockchain techniques to prevent and detect attacks, thus maintaining the security of water networks.      

Projected outcomes

Research results are analysed using water distribution data from one public network and applied to various machining learning techniques to address the lack of labelled data: 

1. Finding a labelling mechanism that can be associated with the attacks indirectly, or;
2. Anomaly detection using unsupervised method. Further smart contract of blockchain technologies are applied to water distribution network to audit data to enhance the CPS security in water infrastructure.

Funding